# Geospatial MCP Server & AI Assistant
A powerful Model Context Protocol (MCP) server for geospatial analysis, featuring a real-time AI Assistant, 3D terrain visualization, and drone mission planning capabilities.
## 🌍 Features
* **Geospatial Tools**:
* `elevation`: Get elevation at any coordinate (uses DTED/SRTM data).
* `line_of_sight`: Calculate visibility between two points over terrain.
* `render_heatmap`: Generate visual elevation heatmaps.
* `distance`: Calculate great-circle distances.
* `get_terrain_grid`: Fetch raw elevation grids for 3D rendering.
* `to_mgrs` / `from_mgrs`: Coordinate conversion.
* **AI Assistant**:
* Chat with your data using open-source LLMs (via Hugging Face).
* Ask questions like "Can I see point B from point A?" or "Plan a drone route."
* Visual responses (heatmaps displayed directly in chat).
* **3D Visualization**:
* Interactive 3D plots of terrain and drone flight paths.
* Support for custom GeoJSON route uploads.
* **MCP Compliant**:
* Implements the Model Context Protocol.
* Can be connected to Claude Desktop or other MCP clients via `bridge.py`.
## 🚀 Getting Started
### Prerequisites
* Docker (recommended) OR Python 3.11+
* A Hugging Face API Token (free) for the AI Assistant.
### Quick Start (Docker)
1. **Clone the repository**:
```bash
git clone <your-repo-url>
cd geo-mcp-server
```
2. **Configure Environment**:
Create a `.env` file in the root directory:
```bash
HF_TOKEN=hf_YourHuggingFaceTokenHere
```
3. **Build and Run**:
```bash
make build
make run
```
4. **Access the Dashboard**:
Open [http://localhost:7860](http://localhost:7860) in your browser.
### Local Development
1. **Install Dependencies**:
```bash
pip install -r requirements.txt
```
2. **Install System Libraries** (if needed for GDAL/Rasterio):
* Ubuntu: `apt-get install gdal-bin libgdal-dev`
* MacOS: `brew install gdal`
3. **Run Server**:
```bash
./start.sh
```
## 📂 Project Structure
* `server.py`: The core FastAPI server implementing the MCP tools.
* `gradio_dashboard.py`: The web UI with AI Assistant and 3D visualization.
* `bridge.py`: An MCP Stdio adapter for connecting to Claude Desktop.
* `dted/`: Directory for storing elevation data (DTED/SRTM files).
* `sample_routes/`: Example GeoJSON files for drone missions.
* `scripts/`: Utilities for downloading data and testing.
## 🤖 AI Assistant
The dashboard includes a "AI Assistant" tab powered by `Qwen/Qwen2.5-72B-Instruct` (via Hugging Face). It can reason about geospatial data and call tools to answer complex queries.
**Example Queries:**
* "What is the elevation at 36.1, -112.1?"
* "Check if a drone at 100m altitude can fly safely from 36.1, -112.1 to 36.15, -112.15."
* "Show me a terrain heatmap around 36.1, -112.1."
## 🔌 Connecting to Claude Desktop
You can use this server as a tool provider for Claude Desktop.
1. Ensure the server is running (e.g., via Docker or locally).
2. Configure Claude Desktop to use `bridge.py` (requires `uv` or python environment):
```json
"geo-server": {
"command": "python",
"args": ["/path/to/geo-mcp-server/bridge.py"]
}
```
*Note: The bridge currently assumes the HTTP server is running at localhost:8000.*
## 🗺️ Data Sources
The server uses DTED (Digital Terrain Elevation Data) or SRTM data.
* Place `.dt2` or `.hgt` files in the `dted/` directory.
* Use `scripts/download_aws_terrain.py` to fetch data from AWS Open Data.
## 📄 License
[MIT License](LICENSE)
